Apr 19, 2005

Local pub gets better response with double opt-in

We’re helping a local chain of pubs (several locations around Atlanta) with their email newsletters. They’ve been collecting email addresses from their patrons for years (from an opt-in registration form on their website, and offline, from their various pubs). One problem they were having is list churn. They had thousands of people on their list, but almost half were hard-bouncing! Plus, their open rates were dismal.

The first thing we did was switch them to double opt-in…

For those of you who aren’t familiar with double (or confirmed) opt-in, it’s a process where someone signs up on your website, then you send them an email asking them to "confirm" their registration by clicking a link first. Only after they click the link are they added as a subscriber. There are lots of benefits to this, such as:

  • Nobody can sign someone else up, without their permission (anybody with a "friend" that signed them up for daily fart jokes will appreciate this precautionary measure)
  • No chance of typos, and erroneous registrations (here’s a famous example)
  • When someone clicks "confirm" you can track their IP address, date, and time of opt-in, and use those records as proof that they registered (in case your competition maliciously reports you for spamming, and your ISP threatens to boot you for it)
  • If they "go through the trouble" of confirming, they must really want to hear from you, so your response rate will be higher

In spite of all the benefits, a lot of marketers are hesitant to implement double opt-in. They’re afraid that there’s too much of a chance that people won’t "go through the trouble of confirming," so their lists won’t grow as fast.

But proponents of double opt-in are quick to argue that those people who actually confirm their registrations are much more likely to respond to your emails, and less likely to opt-out or to report you for spamming (which 13% of recipients do, even if they opted in).

Which is the better route? Quantity or Quality? With the client’s permission, we did a little experimenting…

  • The client wanted to send a 20% off coupon for a Thursday night event at one of their pubs.
  • We segmented the list to only those patrons that expressed interest in that particular pub
  • Next, we divided that list into those that double-opted-in (the "clean list" as we called it) and those that didn’t (this "dirty list" consisted of single-opt-ins and addresses collected offline)
  • The "dirty list" was over 3 times larger than the "clean list" so we were a bit worried at first
  • We sent the e-coupon on a Monday, to give people plenty of time to prepare for the Thursday event

Results from the "Dirty List":

  • Total 16.23% bounce rate, with 2.65% of them hard bounces (undeliverables)
  • 18.54% unique open rate

Results from the "Clean List":

  • Total 1.23% bounce rate, with no undeliverables
  • 45.68% unique open rate

What about actual coupon downloads? What’s really interesting is that even though the "clean list" is 3 times smaller than the dirty list, it got the same number of coupon downloads.

So the "clean list" is cheaper to send to because it’s smaller, it gets a higher open rate, and generated the same number of coupon downloads as the dirty list. If you use another vendor besides MailChimp (shame on you), you probably pay a monthly fee based on your list size (blech), in which case double opt-in can actually save you money just on list management costs.